Tunaman

Lee, Michel -- Just an update. The program that was failing to run on Sunday's over the last 2 weeks did run tonight. Just to recap -- I changed the program by checking all days individually. I also changed the logic so that it runs in a 10 minute window rather than at a specific time.

Hey Michel Grace period was set to 3 minutes -- I would think that would be enough time to clear any run stacks. I have changed it if to use from/to logic, it now will from sunset -10 to sunset - 8 minutes. It did run tonight. It has only failed to run on Sundays, so I guess we'll have to wait until Sunday to see if this works.

It also appears to only be that program. All others that should trigger on Sunday did. I'll set the days individually, and we'll see what happens next week.

Yes, both are 3.1.17.

Lee -- I forgot to mention -- the Status is True

Lee The Last Run Time is Sat 2012/02/11 05:40:19 PM. Last Finish Time is also Sat 2012/02/11 05:40:19 PM. Next Scheduled Run Time is Mon 2012/02/13 05:42:23 PM (seems like it doesn't like Sundays). Screenshot:

This is the second Sunday in a row (last week I thought it might be a fluke) where a program that I have to trigger ~sunset has not run. All Days is selected, and it is set to run daily. It runs all other days of the week, just not on Sundays.

Yes, it is the new InLineLinc, and deleting the scene and re-creating it worked. Thanks!

I downloaded and updated to 3.1.15 -- My 2475SDB is not working correctly. It will respond to a scene where the scene sets the level to 0 (a scene can turn it off), but it will not respond when scene sets the level to 100% (a scene can't turn it on). I think someone else reported that it was working but only after the rebooted (cycled power). I tried that, it didn't correct the problem. About reports both firmware and client at 3.1.15.

Hey Michel. I replaced my M1XEP, and now everything appears to be working. Hopefully, I'll find some time to give it a more through run-through. So far, everything looks good, thanks for your patience while I worked through what turned out to be a non ISY issue.

Michel -- I think I may have found the problem. It looks like my M1XEP is failing. I noticed that the ISY was hanging when it was trying to get the text for the audio devices. I was able to write a quick script which uses socat to ease-drop on the serial port, and I noticed that when the sd command with type 18 was sent, the M1XEP was sending back what looked like a corrupt response (a lot of hex values), however, I don't have any audio devices, so it should have responded with a 000 for the value. I think the ISY then went into a tight loop. I also noticed that when I check the version of the M1XEP software via the vn command, it was returning all 000's. From ElkRP, it shows the correct version. I tried to overwrite it with the same version, but it failed on the upgrade. The interesting part is that everything else that uses the M1EXP (ekeypad for instance) continues to work fine (but nothing is looking for audio devices). I'll reach out to Elk tech supp tomorrow, but based on fact that my experiences seem to be unique, and that I'm seeing some strange responses from the M1EXP, let's wait and see what they have to say.

Michel I got a chance to try 3.1.13. Unfortunately, I'm still seeing same problems. The error log is full of messages: Sat 2011/11/26 10:26:37 PM System -170001 [uDSockets] HTTP:31 error:6

I THINK I did -- it was late last night when I tried. I do recall getting a success message, but I can't recall what the numeric code was.

Michel. I did execute the topology refresh as you suggested. Status is the same -- it does show all of the inputs, keypads, ..., but still no text.

FYI, I did check again. ElkRP is definitely on 2101, and can connect. Plus, the isy is able to track the zone status (I confirmed by opening a door, the ISY did track it. It just can't seem to parse/store the text.

Ok. More info for you. It does eventually populate the zones, keypads and outputs, it just does not have any text in any of the name fields. And it does correctly track the state of all zones, outputs, etc. Still trying to discover the text. I'm going to roll back to 3.1.10 until you have 3.1.13 ready.

ElkRP is set to use port 2101, and does connect successfully.

Michel -- Elk-RP is not currently connected, but I was able to connect last night without issue. For what it is worth, I did make sure to stop all other connections to m1exp last night, and then rebooted the isy. Same behavior.

I can telnet and do see good traffic. Elk firmware is 5.2.8 m1exp is 1.3.28

More diagnostics: Attached gif is system status -- shows 122880b in bad (memory?) status.

I can't physically access my isy right now, but I'm beginning to suspect there may be a hardware/memory issue. I'm guessing that if I could see it, I would also see memory & error lights flashing. I did telnet in to check memory: http://192.168.16.5>sm Static = 0 used, Heap = 3145728 free, Total = 3145728 Cur Cur High Size Tot Free Used Used 600: 1034 983 51 51 197: 4 2 2 2 106: 270 270 0 28 64: 255 249 6 99 132: 90 90 0 0 706: 530 312 218 222 24: 12000 11723 277 277 4096: 10 10 0 2 10240: 5 5 0 0 17408: 5 5 0 0 32: 300 287 13 16 1024: 35 22 13 17 8192: 14 11 3 5 32768: 9 9 0 1 131072: 2 2 0 0 128: 1088 64 1024 1024 8180: 1 1 0 0 4574: 22 17 5 5 While I don't know the inner workings, that looks suspicious to me.

Add me to the few who have experienced problems. Installed 3.1.2, following all directions. Security tab would not show M1 status, so I then installed Elk Module. Things got worse. Now the program folder is not populated. I did notice that it seems to run a 'discovery' process every time the isy is rebooted or whenever I connect the admin console. (I have a home-grown perl data logger so I can see all traffic on the m1 serial interface, thus I can see everything that goes across the m1exp). It seems like it can't save anything (text descriptors), so every time it needs to display any of the elk names (areas, zones, keypads...) it forces a refresh, and the ISY gets very slow.